Streptolydigin-resistant mutants in an evolutionarily conserved region of the beta' subunit of Escherichia coli RNA polymerase.

Mutations conferring streptolydigin resistance onto Escherichia coli RNA polymerase have been found exclusively in the β subunit (Heisler, L. M., Suzuki, H., Landick, R., and Gross, C. A.(1993) J. Biol. Chem. 268, 25369-25375). We report here the isolation of a streptolydigin-resistant mutation in the E. coli rpoC gene, encoding the β‘ subunit. The mutation is the Phe793 → Ser substitution, which occurred in an evolutionarily conserved segment of the β‘ subunit. The homologous segment in the eukaryotic RNA polymerase II largest subunit harbors mutations conferring α-amanitin resistance. Both streptolydigin and α-amanitin are inhibitors of transcription elongation. Thus, the two antibiotics may inhibit transcription in their respective systems by a similar mechanism, despite their very different chemical nature.

Streptolydigin (Stl) 1 is a 3-acyltetramic acid antibiotic (1), which specifically inhibits bacterial DNA-dependent RNA polymerase (2)(3)(4)(5)(6). Stl interacts with RNAP in ternary transcription complexes and inhibits growth of nascent RNA chains during transcription initiation and elongation (4,6). The binding of RNAP to template DNA is not affected by Stl (2). Thus, the likely target of Stl is either the binding of incoming NTP in the substrate binding site of RNAP or the catalysis of phosphodiester bond formation (4,6).
Transcription by RNAP purified from mutant cells that acquired resistance to the drug is resistant to Stl (3). In Escherichia coli, all Stl-resistant RNAPs studied to date have an altered ␤ subunit (7,8), and all known mutations leading to Stl R RNAP map to the rpoB gene, which codes for the ␤ subunit (8). Substitutions of amino acids in ␤ between 540 and 546 lead to Stl resistance (9 -11). The highest resistance levels in vivo and in vitro were found in the case of substitutions at positions 544 and 545 (10 -11). In the absence of more direct data it has been assumed that these ␤ amino acids participate in Stl binding to RNA polymerase.
Since Stl inhibits elongation of nascent RNA, mutations changing RNAP Stl-binding site are likely to change the catalytic properties of the enzyme. However, RNAP with the known ␤ subunit Stl R mutations has unaltered transcription elongation and transcription termination properties in vivo and in vitro (11). Moreover, the site of Stl resistance in the ␤ subunit is dispensable for RNAP function, since mutant RNAPs with deletions spanning amino acids 534 -545 are functional both in vivo and in vitro (10). This apparent discrepancy led Heisler et al. (11) to hypothesize that other site(s) in RNAP may be involved in Stl binding. This hypothesis was substantiated by a report (12) that in RNAP reconstituted in vitro, the ␤Ј subunit from a mutant strain of Bacillus subtilis was responsible for Stl resistance. However, no ␤Ј subunit Stl R mutants have ever been reported in E. coli. Here, we report an isolation and localization of such a mutation. The mutation leads to an amino acid substitution in an evolutionarily conserved region of the ␤Ј subunit. In the ␤Ј homologues from eukaryotic RNA polymerase II, this region harbors mutations that lead to resistance to ␣-amanitin, a peptide toxin that specifically inhibits RNA chain elongation by RNA polymerase II. Our results raise the possibility that the structurally different inhibitors streptolydigin and ␣-amanitin may interact and inhibit RNA polymerase from prokaryotic and eukaryotic systems by a similar mechanism.

EXPERIMENTAL PROCEDURES
Bacterial Techniques and DNA Manipulations-Streptolydigin was purchased from BioMol. Cells were grown in LB medium. Plates containing linear gradient of Stl concentration were prepared as described (10). To derepress the lac promoter of the pMK201 plasmid 1 mM isopropyl-1-thio-␤-D-galactopyranoside was included in the selective media. In vitro hydroxylamine mutagenesis of the pMK201 plasmid was carried out exactly as described by Heisler et al. (11). Standard laboratory techniques were used for subcloning and sequencing of rpoC fragments.
Preparation of Mutant RNA Polymerases and in Vitro Transcription-To purify histidine-tagged RNAP from induced cells containing pMK201 or its derivatives, the procedure described by Kashlev et al. (13) employing affinity chromatography on Ni 2ϩ -NTA-agarose (Qiagen) was used.
Transcription of bacteriophage T2 DNA was performed in 100-l reactions containing 10 g of T2 DNA, 2 g of WT or mutant RNAP, 0.5 mM ATP, CTP, and GTP, 0.025 mM [ 3 H]UTP (43 Ci/mM), 10 mM Tris-HCl (pH 7.9), 100 mM NaCl, 50 mM KCl, 10 mM MgCl 2 , 0.5 mg/ml bovine serum albumin, and different concentrations of Stl. Reactions were initiated by addition of NTPs and proceeded for 15 min at 37°C. Reactions were terminated by addition of 1 ml of 10% trichloroacetic acid, and the amount of acid-insoluble radioactivity was determined.
To determine the elongation rate of the mutant RNAP, elongation * This work was supported in part by grants from the Lucille P. Markey Charitable Trust, the Irma T. Hirschl Trust, and the Human Reactions proceeded for 15 min at 23°C and were washed three times with 1.5 ml of the buffer (40 mM Tris-HCl (pH 7.9), 40 mM KCl, 10 mM MgCl 2 ) as described (13). ϩ20 elongation complexes were synchronously started by making reactions 10 M with NTPs. Reactions proceeded for 0 -120 s at 23°C. Reactions were terminated by addition of EDTA to 15 mM. To determine transcription termination efficiencies of mutant enzymes, ϩ20 elongation complexes were prepared as described above using 324-bp DNA fragment (template 1 of Nudler et al.; Ref. 39) containing T7 A1 promoter followed by phage tR2 terminator. Transcription was started by addition of 1 mM NTPs. Reactions proceeded for 2 min at 23°C. Products were analyzed by urea-polyacrylamide gel electrophoresis (7 M urea, 6% polyacrylamide), followed by autoradiography and PhosphorImager analysis.

RESULTS
The purpose of this study was to generate mutations in E. coli rpoC that would result in Stl R RNAP. E. coli cells are naturally resistant to high concentration of Stl, due to a permeation barrier (2). Mutant E. coli that become sensitive to low concentration of Stl can be used for selection of mutants with Stl-resistant RNAP (2,11). Throughout this work, we used the Stl-sensitive E. coli strain CAG 14064, provided by C. Gross. To obtain Stl R mutations, rpoC expression plasmid pMKa201 (13) was mutagenized with hydroxylamine in vitro (11). After mutagenesis, plasmid DNA was introduced into CAG 14064 strain by electroporation and cells were plated on media containing 12.5 g/ml Stl. Out of ϳ1 ϫ 10 4 plasmid-bearing cells plated on the selective media, 20 Stl R clones were obtained. Plasmid DNA was prepared from the resistant clones and retransformed in CAG 14064, and cells were plated on Stl-containing media. In this way two pMKa201 derivatives (pMKa201-15 and pMKa201-16) that conferred Stl resistance to CAG 14064 strain were selected and used for further analysis.
To localize Stl R mutations, a series of in vitro exchanges of DNA fragments between the mutant plasmids pMKa201-15 and pMKa201-16, and the parental pMKa201 were performed. The recombinant plasmids were checked for their ability to support growth of CAG 14064 cells on 12.5 g/ml Stl. In this way, the determinant of the Stl R phenotype in both mutant plasmids was shown to reside in the 997-bp SphI-SalI fragment of rpoC (Fig. 1A). The fragment was sequenced in both orientations in pMKa201-15, pMKa201-16, and the parental pMKa201. Comparison of the sequences revealed a single nucleotide difference between pMKa201 and the mutant plasmids. Both mutant plasmids have a TTC at codon position 793, while pMKa201 has a CTC at that codon position, as does the published sequence (14). As a result of this change, the ␤Ј subunit expressed from the mutant plasmids has a phenylalanine at amino acid position 793 instead of serine, as found in the wild-type protein (Fig. 1B). Since we have sequenced the entire fragment responsible for Stl resistance of pMKa201-15 and pMKa201-16 and the CG to TA transition is the only change from the wild-type sequence, we conclude that the change from phenylalanine to serine at ␤Ј position 793 is the cause of Stl R phenotype. We have named the Stl R mutation rpoCS793F and will refer to the mutation by that name. The results of plating of CAG 14064 cells expressing rpoCS793F from the pMKa201 plasmid on a plate containing a linear gradient of Stl are shown in Fig. 2A. Cells overproducing the mutant ␤Ј subunit continued to grow at ϳ20 g/ml Stl, while cells overproducing wild-type ␤Ј from the pMKa201 failed completely to form colonies on the gradient plate.
The ␤Ј subunit expressed from the plasmid pMKa201 or its derivatives is extended with a stretch of six consecutive histidine residues at its C terminus (the His tag). As is shown elsewhere, the His-tagged RNAP is indistinguishable from the wild-type RNAP in functional tests and can be easily separated from RNAP with chromosome-encoded ␤Ј by affinity chromatography on Ni 2ϩ sorbent (13). We purified His-tagged RNAP from cells harboring pMK201 or pMK201-rpoCS793F.The response of the two enzymes to Stl was compared in bacteriophage T2 DNA transcription assay (Fig. 2B). The enzymes displayed equal levels of activity in the absence of Stl (data not shown). In the presence of Stl, the mutant enzyme was clearly more active than the control WT enzyme (half-inhibition at 100 and 10 g/ml Stl, respectively).
Recently, one of us performed a systematic search for termination-altering mutations in the cloned E. coli rpoC gene (15). The mutations clustered in several regions of the gene. Many of the termination-altering mutations resulted in amino acid substitutions in a segment of the ␤Ј subunit between amino acids 630 and 800 (interval 3, see Ref. 15 for nomenclature). Since rpoCS793F is contained within interval 3, we investigated the ability of interval 3 termination-altering mutations to confer Stl R phenotype to Stl-sensitive cells. CAG 14064 cells harboring 19 pRW308 rpoC expression plasmid derivatives carrying interval 3 mutations were streaked on plates with a linear gradient of Stl. As a control, 8 interval 2 (amino acids 310 -390) and 10 interval 5 (amino acids 1305-1370) mutant plasmids were used. Out of 37 plasmids tested, only 5 plasmids with interval 3 mutations (rpoC3302 (M747I), rpoC3309 (R780H), rpoC3310 (G729D), rpoC3312 (E756K), and rpoC3329 (M725I)) conferred very low levels of Stl resistance to CAG 14064 cells (Fig. 3A, and data not shown). rpoC alleles 3302 (M747I) and 3309 (R780H), which conferred higher levels of resistance, were recloned in the pMKa201 plasmid; the two His-tagged mutant RNAPs were purified, and their response to Stl was investigated in the T2 DNA transcription system (Fig. 3B). In the absence of Stl, the mutant enzymes were 50% more active than the control enzyme (data not shown). The two enzymes reproducibly demonstrated slightly higher levels of Stl resistance than the control enzyme (half-inhibition at 25 g/ml Stl).
Since Stl inhibits phosphodiester bond formation, it is expected that mutations in Stl-binding site will change the catalytic properties of RNAP. Transcription elongation, transcription pausing, and transcription termination by the three mutant enzymes (M747I, R780H, and S793F) and WT RNAP were investigated in the experiment presented in Fig. 4. The three mutant enzymes elongated RNA at different rates (Fig.  4A); S793F RNAP was slightly "slower" than the WT enzyme, while M747I and R780H enzymes were considerably "faster," in agreement with the previous data (15). All three mutant enzymes and the WT control demonstrated essentially the same pausing pattern in this assay. Changes in transcription elongation rates were accompanied by changes in transcription termination efficiencies of the mutant enzymes on a factorindependent tR2 terminator (Fig. 4B); S793F RNAP terminated slightly more efficiently (10% read-through), while M747I and R780H enzymes terminated considerably less efficiently (53 and 48% read-through, respectively) than the wildtype enzyme (20% read-through).
The response of the mutant enzymes to other transcription inhibitors also was investigated. The mutant enzymes were as sensitive to initiation inhibitor rifampicin and elongation inhibitor tagetitoxin as the wild-type control (data not shown).

CONCLUSIONS
The principal result of this work is the demonstration that mutations in the ␤Ј subunit of E. coli RNAP can confer resistance to Stl. From the point of practical E. coli RNAP genetics, the availability of an Stl resistance marker in rpoC should facilitate isolation of loss-of-function rpoC mutations, similarly to the approach that was used with rpoB (RNAP ␤ subunit) mutations employing the rifampicin resistance marker (16,17).
Biochemical analysis of the three mutant RNAPs (M747I, R780H, and S793F) demonstrates that the extent of defects in transcription elongation, transcription pausing, and transcription termination of the mutants studied is not correlated with the levels of Stl resistance. This situation is reminiscent of that for the ␤ subunit Stl R mutants, which do not demonstrate significant transcription defects (11). We note that the S793F mutation occurred in the highly conserved Segment F of ␤Ј (Fig. 1B), and that Stl R mutations affecting RNAP basic function may have escaped our screen which requires mutant RNAP in vivo function. It is conceivable that additional changes close to E. coli ␤Ј position 793 will be identified that will lead to higher levels of streptolydigin resistance. While this work was in progress, the B. subtilis Stl R rpoC mutation (12) was sequenced (40). The B. subtilis mutation projects on E. coli ␤Ј position 792, i.e. just next to S793F mutation isolated in this study. Thus, the ␤Ј determinants of RNAP streptolydigin resistance coincide in Gram-negative E. coli and Gram-positive B. subtilis.
Bacterial RNAP ␤Ј subunits are highly homologous to the large subunit of eukaryotic RNA polymerase II (18). As is shown in Fig. 1B, the ␤Ј Stl R mutations characterized in this work occurred in a segment of ␤Ј that is highly conserved in evolution (Segment F, according to Ref. 15 nomenclature). Several lines of evidence suggest that this segment plays an important role conserved in all RNA polymerases. Mutations in Segment F that dramatically change nascent RNA elongation rate and/or termination efficiencies were reported in eukaryotic as well as prokaryotic RNAPs (19,15). A recent cross-linking study demonstrates that the E. coli ␤Ј segment between amino acids 748 and 814, containing most of segment F, is in tight contact with the 3Ј end of the nascent RNA. 2 Finally, in eukaryotic RNA polymerase II largest subunit, Segment F harbors mutations that render RNA polymerase resistant to the elongation inhibitor ␣-amanitin (Fig. 1B). The discovery of Stl R mutations in Segment F suggests that despite the lack of structural similarity, Stl and ␣-amanitin may inhibit transcription by a similar mechanism. Although speculative, this hypothesis is consistent with available biochemical data on the mechanism of Stl and ␣-amanitin inhibition of transcription. (i) Both are elongation inhibitors, but allow several phosphodiester bonds to be made and different complexes are inhibited to a different extent (4,20,21); (ii) both inhibit pyrophosphorolysis 3,4 (2); (iii) both inhibit nascent RNA cleavage by transcription elongation factors (22,23). 4